喻园管理论坛2023年第46期(总第869期)
演讲主题: DynamicData-DrivenSimulation: Real-Time Data for Dynamic System Analysis and Prediction
主 讲 人: 胡晓琳 Hu Xiaolin，美国乔治亚州立大学计算机科学系教授
主 持 人: 胡斌，管理学院管理科学与信息管理系教授
活动时间: 2023年6月5日（周一）9:00-10:30
活动地点: 管理大楼119教室
主讲人简介：
Xiaolin Hu博士是美国乔治亚州立大学计算机科学系的教授，他博士毕业于美国亚利桑那大学，研究领域为计算机模拟理论与应用、复杂系统科学、多智能体建模与模拟、并行计算、云计算，他将这些理论与方法应用于生态科学、行为与社会科学、公共健康与管理等领域。主持了5项与大规模实时模拟、群体行为模拟有关的美国科学基金项目（NSF），是NSF CAREER Award获奖者。担任了数个建模与仿真领域中的国际会议/研讨会的学术委员会主席，并担任如下期刊的副编辑：SimulationTransaction of the Society for Modeling and Simulation，International Journal of Modeling、Scientific Computing，International Journal of Agent Technologies and Systems。已发表80多篇学术论文，担任NSF（美国科学基金会）和NIH（美国国立卫生研究院）的评审委员。
活动简介：
The increasing amount of data collected from dynamic systems in real time poses a fundamental question to the modeling and simulation community: how to integrate real-time data with simulation to enable real-time prediction and analysis for a dynamic system under study? Answering this question requires a systematic examination of the relationship between data and simulation, as well as new ways of making (real-time) data and simulation work together. This talk will presentDynamic Data Driven Simulation (DDDS) as a new simulation paradigm that combines real-time data with simulation models for studying dynamic systems.A set of DDDS activities will be described, including dynamic state estimation, online model calibration, external input forecasting, and real-time simulation-based prediction/analysis. Their roles in DDDS and their relationships with other modeling and simulation activities will be discussed. The talk will also introduce data assimilation as an enabling technique for DDDS and shows how data assimilation can be carried out for discrete simulations.
